Kidney Week

Abstract: TH-PO1244

Pentoxifylline for Diabetic Kidney Disease: A Systematic Review, Meta-Analysis, and Trial Sequential Analysis

Session Information

• Late-Breaking Research Posters
  October 22, 2026 | Location: Exhibit Hall A, Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Cardiovascular-Kidney-Metabolic Health

• 602 Cardiovascular-Kidney-Metabolic Health: Clinical

Authors

• Agarwal, Rajiv, Richard L Roudebush VA Medical Center, Indianapolis, Indiana, United States

Rajiv Agarwal, MD, MS, FASN

• Leehey, David J., Edward Hines Junior VA Hospital Cooperative Studies Program, Hines, Illinois, United States

David J. Leehey, MD

• Navaneethan, Sankar D., Michael E DeBakey VA Medical Center, Houston, Texas, United States

Sankar D. Navaneethan, MD, MS, MPH, FASN

• Utzschneider, Kristina M., VA Puget Sound Health Care, Seattle, Washington, United States

Kristina M. Utzschneider, MD

• Xing, Bin, Edward Hines Junior VA Hospital Cooperative Studies Program, Hines, Illinois, United States

Bin Xing, PhD

• Carlson, Kimberly, Edward Hines Junior VA Hospital Cooperative Studies Program, Hines, Illinois, United States

Kimberly Carlson, MS

Background

Pentoxifylline is an inexpensive oral anti-inflammatory agent tested in diabetic kidney disease (DKD) for four decades. Prior meta-analyses report reduced albuminuria, but none reports a prediction interval (PI) or asks whether enough evidence has accrued.

Methods

Systematic review, meta-analysis, and trial sequential analysis (TSA) of randomized trials of oral pentoxifylline in DKD; 7 databases searched to June 2026 without language restriction (PROSPERO CRD420261413599). Co-primary outcomes: albuminuria (ratio of means, RoM) and GFR change (mean difference, MD); primary hard outcome: end-stage kidney disease (ESKD) or death (risk ratio, RR). Random-effects models; risk of bias by RoB 2; a 95% PI and TSA accompany every estimate.

Results

Of 34 included studies, 32 contributed data. Albuminuria fell 31% (RoM 0.69, 95% CI 0.50-0.94; 22 studies contributing 23 comparisons), but I2 was 97%, the PI 0.15-3.10, and 53% of the required information had accrued. GFR favored pentoxifylline (MD 4.49 mL/min/1.73 m2, 0.84-8.14; 5 studies; 11% accrued), but was 2.96 (-15.34 to 21.26) in the 2 comparisons at low or some-concerns risk of bias. For ESKD or death, 29 events in 6 studies contributing 7 comparisons gave RR 0.90 (0.46-1.77), 3% of that required. Gastrointestinal adverse events were tripled (RR 3.00, 1.94-4.64); serious adverse events were not (0.83, 0.56-1.22). No monitoring boundary was crossed for any outcome.

Conclusion

Mean effects are estimable but fragile: every prediction interval crosses no effect, and the albuminuria estimate depends on which trials are counted. Kidney failure and death have scarcely been tested: 3% of the required evidence. Gastrointestinal intolerance is the firmest finding. An adequately powered, long-term trial reporting hard endpoints is the priority.

Acknowledgment

Funding: None. This review had no funding and no support from any academic institution or commercial sponsor; it was not funded by the VA Cooperative Studies Program.
Disclosure: All six authors participate in the VA Cooperative Studies Program pentoxifylline trial (CSP #2008), an ongoing pragmatic randomized trial of pentoxifylline in diabetic kidney disease (RA Co-PI, DJL PI).

Funding

• Veterans Affairs Support